Advanced

Harmonic minimum mean squared error filters for multichannel speech enhancement

Jensen, Jesper Rindom ; Christensen, Mads Groesboll and Jakobsson, Andreas LU (2017) 42nd IEEE International Conference on Audio, Speech, and Signals Processing, ICASSP 2017 p.501-505
Abstract

Many state-of-the-art multichannel speech enhancement methods rely on second-order statistics of the desired speech signal, the noise signal, or both. Estimation of those are difficult in practice, resulting in a practical performance that is typically much lower than their potential theoretical performance. We propose two multichannel enhancement techniques that instead rely on a model for voiced speech. That is, the proposed methods are driven by the signals' fundamental frequencies, which may be accurately estimated even in noisy scenarios. The first method is designed independently of the microphone array geometry and source position, whereas these are utilized in the second approach. Thereby, we can investigate when to exploit such... (More)

Many state-of-the-art multichannel speech enhancement methods rely on second-order statistics of the desired speech signal, the noise signal, or both. Estimation of those are difficult in practice, resulting in a practical performance that is typically much lower than their potential theoretical performance. We propose two multichannel enhancement techniques that instead rely on a model for voiced speech. That is, the proposed methods are driven by the signals' fundamental frequencies, which may be accurately estimated even in noisy scenarios. The first method is designed independently of the microphone array geometry and source position, whereas these are utilized in the second approach. Thereby, we can investigate when to exploit such information in the case of localization errors and violations of the spatial assumptions. Numerical results show that the proposed method is able to outperform competing methods in terms of both output SNRs and PESQ scores.

(Less)
Please use this url to cite or link to this publication:
author
organization
publishing date
type
Chapter in Book/Report/Conference proceeding
publication status
published
subject
keywords
multichannel speech enhancement, voiced speech, MMSE filtering, harmonic filters, DOA mismatch
host publication
2017 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2017 - Proceedings
article number
7952206
pages
5 pages
publisher
Institute of Electrical and Electronics Engineers Inc.
conference name
42nd IEEE International Conference on Audio, Speech, and Signals Processing, ICASSP 2017
conference location
New Orleans, United States
conference dates
2017-03-05 - 2017-03-09
external identifiers
  • scopus:85023776452
  • scopus:85023776452
ISBN
9781509041176
DOI
10.1109/ICASSP.2017.7952206
language
English
LU publication?
yes
id
a0558000-3137-45e0-bd8e-63630c3cf8b0
date added to LUP
2017-02-14 10:32:22
date last changed
2020-01-16 02:41:06
@inproceedings{a0558000-3137-45e0-bd8e-63630c3cf8b0,
  abstract     = {<p>Many state-of-the-art multichannel speech enhancement methods rely on second-order statistics of the desired speech signal, the noise signal, or both. Estimation of those are difficult in practice, resulting in a practical performance that is typically much lower than their potential theoretical performance. We propose two multichannel enhancement techniques that instead rely on a model for voiced speech. That is, the proposed methods are driven by the signals' fundamental frequencies, which may be accurately estimated even in noisy scenarios. The first method is designed independently of the microphone array geometry and source position, whereas these are utilized in the second approach. Thereby, we can investigate when to exploit such information in the case of localization errors and violations of the spatial assumptions. Numerical results show that the proposed method is able to outperform competing methods in terms of both output SNRs and PESQ scores.</p>},
  author       = {Jensen, Jesper Rindom and Christensen, Mads Groesboll and Jakobsson, Andreas},
  booktitle    = {2017 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2017 - Proceedings},
  isbn         = {9781509041176},
  language     = {eng},
  month        = {06},
  pages        = {501--505},
  publisher    = {Institute of Electrical and Electronics Engineers Inc.},
  title        = {Harmonic minimum mean squared error filters for multichannel speech enhancement},
  url          = {http://dx.doi.org/10.1109/ICASSP.2017.7952206},
  doi          = {10.1109/ICASSP.2017.7952206},
  year         = {2017},
}